Skip to main content

Table 6 Classification results of the proposed technique compared to simple feature fusion with four datasets using F-KNN, ES-KNN and SVM

From: A multilevel features selection framework for skin lesion classification

Vector FusionOA (%)
Feature Fusion ApproachProposed (ECNCA)
F-KNNSVMES-KNNF-KNNSVMES-KNN
\(\hbox {PH}^{2}\)
FV0-FV182.880.080.196.993.795.1
FV0-FV282.181.781.795.194.893.2
FV1-FV282.982.082.197.495.097.1
FV0-FV1-FV283.282.282.4\(98.8*\)95.198.1
ISIC-MSK
FV0-FV174.271.774.693.787.288.8
FV0-FV273.971.073.089.189.087.4
FV1-FV273.172.575.186.591.089.7
FV0-FV1-FV276.474.874.9\(99.2*\)95.196.9
ISIC-UDA
FV0-FV171.970.075.988.880.184.7
FV0-FV273.371.274.190.784.588.0
FV1-FV274.175.975.892.882.794.2
FV0-FV1-FV276.573.576.0\(97.1*\)93.395.7
ISBI-2017
FV0-FV173.270.971.188.588.088.8
FV0-FV274.770.772.889.787.389.3
FV1-FV272.170.573.390.088.990.7
FV0-FV1-FV275.375.176.194.193.4\(95.9*\)
  1. * Shows the highest value in each dataset